Conventional bright-field electron tomographic tilt series are obtained by collecting electrons that have traversed a specimen illuminated by a broad beam. Using this approach, the thickness is limited by the severe image blurring that occurs when electrons that have undergone multiple energy losses are focused by the objective lens of the microscope. Furthermore, the maximum area of the image is limited by the depth-of-field of the objective lens, so that only part of the sample is in focus at high tilt angles. Tomographic reconstruction using STEM with a tightly focused electron probe can overcome some of the limitations imposed by tomographic reconstruction using conventional TEM. First, because the incident STEM probe can be focused at any point in a specimen, large areas are imaged in focus even for high tilt angles. Second, because in STEM there are no image-forming lenses after the specimen, the resolution attainable in images of thick specimens is not further degraded by electrons that have suffered multiple energy losses. The most commonly applied STEM approach makes use of an annular dark-field detector to collect electrons that are scattered to high angles. However, the dark-field STEM technique is not well-suited to imaging thick biological specimens because of the limited depth of field defined by the large convergence angle of the incident electron probe. A tenfold or higher increase in depth of field is possible by adjusting the microscope optics to decrease the convergence semi-angle to approximately 1 mrad. Another limiting feature of dark-field STEM as applied to imaging thick specimens is the severe degradation in spatial resolution that occurs toward the bottom surface of a section because of beam broadening. In contrast, we found that much higher spatial resolution can be obtained by collecting only those electrons that are scattered to low angles, that is, by using an axial bright-field detector. Electrons that undergo multiple elastic scattering are substantially displaced from the point of incidence of the STEM probe and have, on average, larger net scattering angles. A large fraction of these electrons can thus be excluded from images recorded with an axial detector, leading to an improvement in spatial resolution toward the bottom surface of thick specimens. We quantified this unexpected improvement in resolution using Monte Carlo electron-trajectory simulations. In order to improve the quality of STEM tomographic reconstructions, we have investigated the mathematically based compressed sensing technique, which makes use of known sparsity in structures of interest. We found that compressed sensing reconstructions outperformed the most commonly used weighted back projection methods in full and undersampled tomogram recovery. However, the advantage of compressed sensing when applied to complex biological structure was less than when the technique is applied to reconstruct simpler inorganic structures such as nanoparticles. We propose that this disparity stems from the increased structural complexity of biological systems, as quantified by dataset sparsity in multiple transform domains. This proposition was supported by numerical simulations using phantom datasets. The numerical nonlinear decoding process of compressed sensing shares strong connections with popular regularized least-squares methods, and the use of such numerical recovery techniques for mitigating artifacts and denoising in reconstructions of fully sampled datasets remains advantageous. We have applied axial (bright-field) STEM tomography to investigate the structure of synaptic ribbons, which are presynaptic protein structures found at many synapses that convey graded, analog sensory signals in the visual, auditory, and vestibular pathways. Ribbons, typically anchored to the presynaptic membrane and surrounded by tethered synaptic vesicles, are thought to regulate or facilitate vesicle delivery to the presynaptic membrane. No direct evidence exists, however, to indicate how vesicles interact with the ribbon or, once attached, move along the ribbons surface to reach the presynaptic release sites at its base. To address these questions, we have tested a passive vesicle diffusion model of retinal rod bipolar cell ribbon synapses. STEM tomography gave us 3D structures of rat rod bipolar cell terminals in 1-micrometer thick sections of retinal tissue at an isotropic spatial resolution of approximately 3 nm. The resulting structures were then incorporated, along with previously published estimates of vesicle diffusion dynamics, into numerical simulations that accurately reproduced electrophysiologically measured vesicle release/replenishment rates and vesicle pool sizes. The simulations suggest that, under physiologically realistic conditions, diffusion of vesicles crowded on the ribbon surface gives rise to a flow field that enhances delivery of vesicles to the presynaptic membrane without requiring an active transport mechanism. We have applied the technique to visuallize synaptic spines in cultured slices of rat hippocampus. It has been possible for the first time to visualize entire post-synaptic densities and to assess differences in ultrastructure that occur when certain important proteins such as PSD-95 are knocked down. In other experiments, it has been feasible to characterize entire ribbon synapses in rat retina and to visualize the precise organization of secretory vesicles within those structures. We have also applied the technique to investigate the association of membranes with the mother centriole in the early stages of generation of primary cilia, which play essential roles in signal transduction. Defects in cilium formation or function cause ciliopathies. The advantage of STEM tomography in this study is its ability to provide 3D reconstructions of the entire centriole assemby. In another application, we have applied STEM tomography to determine the three-dimensional arrangement of the complex membrane systems that are present in blood platelets, including the open canalicular system, dense tubules, and secretory granules. Collecting STEM tomograms from whole platelets provides a new way to visualize the structural changes that occur on activation of platelets that have been prepared by rapid freezing and freeze-substitution. Thus we have demonstrated the feasibility and advantages of STEM using axial detection for imaging thick sections at a spatial resolution around 5-10 nm, which is comparrable to the spatial resolution of conventional electron tomography from thinner sections (typically 3 to 8 nm). Most modern electron microscopes can be operated in STEM mode and can be readily equipped with a bright-field detector, which is anticipated to facilitate implementation of the technique. The demand for high-resolution, large-volume imaging of biological specimens has been addressed so far by the large-scale application of conventional electron tomography of thin sections. Our current work suggests that it will be possible to reconstruct intact organelles, intracellular pathogens and even entire mammalian cells through serial thick-section tomography.